Support

Google Trends doesn’t publish search counts. It scales each chart so its highest point is 100, and every other value is relative to that peak. 50 means half as popular as the peak; 0 means too little data.

So values from two separate charts aren’t directly comparable. To compare terms, set Comparing terms to Compare terms on one scale.

Google only shows data when enough people searched for a term in that place and time range. Very niche terms, small regions and short time ranges often have none. You get a free status row with status: "no_data" instead.

Try a longer time range, a bigger location, or a broader term. Related topics are often empty from Google at the moment (checked 14 November 2026); that’s Google’s choice, not a fault in the run.

Google compares at most 5 terms at a time. The Actor splits your terms into groups of 4 and adds the same anchor term to every group. Because the anchor appears everywhere, its level in each group shows how that group’s scale relates to the first group. The Actor rescales each group by that ratio, then sets the overall peak to 100.

These values are estimates, marked isEstimate: true, with the original value in rawValue. Google rounds to whole numbers, so precision drops if the anchor is tiny next to your other terms. Pick an anchor with mid-range popularity. If the anchor has no data in a group, those values stay on their own scale and you get a warning row.

The last point of an interest-over-time chart usually covers a period that isn’t finished yet, such as the current week. Google marks it as partial and the Actor sets isPartial: true. Its value can still change, so leave it out of trend calculations or refresh it later.

Google limits how fast anyone can fetch Trends data. When it refuses a request, the Actor waits with exponential backoff and random jitter, switches to a fresh proxy session, and tries again, up to the Retries per request setting (8 by default). If Google blocks a request twice on the main proxy, it moves to Apify residential proxy, at no extra cost to you. Anything still missing gets one more retry pass at the end of the run.

If a term still fails, you keep every other result, and a free status row says what happened. If you see many failures, lower Parallel requests and leave Fall back to residential proxy turned on.

Click Abort on the run’s page in Apify Console, or call the Apify API’s abort endpoint. The Actor finishes the step it’s on, saves its progress and stops. Everything already saved stays in your dataset, and you’re only charged for results that were saved.

To limit spending before you start, set a maximum charge for the run. The Actor stops when it reaches it and tells you how many rows it saved.

Apify bills for the Actor, so charges and refunds go through Apify. If a charge looks wrong, open an issue on the Actor’s Issues tab with the run link, or contact Apify support. We’ll look at the run and tell you what we find.

You’re never charged for terms with no data or that failed: those status rows are free.

No. Google publishes trending searches per country, not worldwide. Add each country you want, such as GB or US, to Locations.

Yes. Connect your agent to Apify’s MCP server at https://mcp.apify.com and add this Actor as a tool. The agent can then run it with search terms and read the results from the dataset.